Abstract Breast cancer is one of the leading causes of cancer-related mortality among women worldwide, and the overwhelming majority of breast cancer patient deaths are caused by metastatic disease rather than primary tumor growth. However, the complex process of tumor cell dissemination and colonization of distant tissues is incompletely understood and is largely incurable using existing therapies. Metastatic disease is particularly intractable because metastasis is not solely driven by tumor cell-intrinsic properties but is instead a consequence of dynamic, heterotypic interactions between cancer cells and other cell types in the tumor microenvironment (TME) including cancer-associated fibroblasts (CAFs). CAFs are a prominent component of the breast tumor microenvironment and are important during multiple stages of tumor development, including metastasis. However, CAFs have been shown to display paradoxical functions across different cancer types, suggesting that heterogeneous CAF subpopulations may play diverse or opposing roles in the tumor microenvironment. It was recently shown that periostin, a TGF?-induced matricellular protein that serves as a protein scaffold to promote collagen cross-linking and extracellular matrix stiffening, can distinguish CAF subtypes in breast cancer. This specific CAF subpopulation is not well characterized, and its source, functional properties, and contribution to disease progression have not been studied in vivo. Thus, I propose to use in vivo genetic labelling/targeting of periostin+ subpopulations and their progeny to better delineate their expansion and function during breast tumor growth and metastasis. My preliminary data indicates that ZSGreen-labelled periostin-expressing CAFs are more abundant in the primary tumors and metastases of high-metastatic breast tumor cells compared to their low-metastatic counterparts. I also observed that the collagen organization of high-metastatic tumors containing abundant periostin-expressing CAFs is altered, with high-metastatic tumors displaying long, aligned collagen fibers and low-metastatic tumors displaying disorganized collagen. Accordingly, I found that knocking down periostin in primary human CAFs inhibits tumor cell invasion through collagen in a CAF/tumor cell co-culture spheroid assay, indicating that periostin- expressing CAFs may remodel the extracellular matrix via collagen cross-linking, making it more conducive to cell migration and invasion. In addition, I observed that ZSGreen+ cells in highly metastatic tumors are larger and more spindle-shaped, and that periostin knockdown in primary CAFs reduces their cellular area and ability to migrate, suggesting that periostin is associated with biomechanical properties of CAFs that function in cell spreading and motility. I propose that high-metastatic cancer cells have an enhanced ability to activate periostin-expressing CAFs, driving their proliferation and expansion in the TME and subsequently promoting collagen remodeling and collective cell migration of CAFs and cancer cells, resulting in metastasis.